Evaluating Cost Structure and Economies of Scale of Beef Cattle Fattening … · research has been carried out on dairy farming. In recent years, given the rising production costs

J. Agr. Sci. Tech. (2019) Vol. 21(Suppl.): 1753-1766

1753

1 Department of Agricultural Economics, Faculty of Agriculture, Ferdowsi University of Mashhad,

Islamic Republic of Iran. 2 Department of Agricultural Economics, University of Kentucky, Kentucky, U. S. A. 3 Department of Management, Faculty of Economics and Administrative Sciences, Ferdowsi University of

Mashhad, Islamic Republic of Iran. *Corresponding author; e-mail: [email protected]

Evaluating Cost Structure and Economies of Scale of Beef

Cattle Fattening Farms in Mashhad City

P. Alizadeh1, H. Mohammadi1*, N. Shahnoushi1, S. Saghaian2, and A. Pooya3

ABSTRACT

In recent years, the high cost of raising livestock and, consequently, the sharp increase in

the price of red meat in Iran have reduced its demand, and people consume chicken meat

as a substitute for it. This has reduced the production incentives and, with the bankruptcy

of some beef cattle farms, the welfare of producers and consumers of this product face

serious danger. To overcome this problem, understanding cost structure and reducing

consumer price by reducing production costs seems necessary. Therefore, the aim of this

research was to evaluate cost structure and economies of scale of beef cattle farms in

Mashhad. For this purpose, the short-run Translog cost function along with input cost

share equations were estimated using the iterated seemingly unrelated regression method.

The data were collected in 2017 from beef cattle producers by interview using structured

questionnaires. The result showed that there were increasing returns to scale for all

farms. In addition, the demands for all inputs were perfectly inelastic. On the other hand,

there was weak complementary and substitute relationship between inputs. According to

the results of this research, the most important factor of beef production in the selected

farms was feed, whose demand was inelastic and the possibility of substituting it with

other inputs was very weak. Therefore, the adoption of policies by the government,

including subsidies for feeding cattle and increasing the import of this input, can reduce

the production cost and prevent beef prices from rising.

Keywords: Feeding cattle, Input demand, Iterated seemingly unrelated regression,

Substitution elasticity, Translog cost function.

INTRODUCTION

Although food consumption patterns have

changed over time, meat is still an important

meal component for consumers (Grunert,

2006). Meat is rich in proteins, vitamins,

minerals, micronutrients and fat. It is

necessary for health and is one of the main

components of human eating habits (Iran

Ministry of Agriculture Jihad, 2015). The

main source of meat production is livestock.

About 45 percent of the value added of

agriculture sector in Iran is related to animal

husbandry and about 3 million people are

directly involved in animal husbandry

(Fatemi Amin and Mortezaei, 2013). The

consumption of red meat in Iran, especially

in rural areas and in low-income groups, is

low compared to developed countries

(Rahimi Baigi et al., 2014). According to the

global standard, the per capita consumption

of red meat is about 30 to 45 kg (FAO,

2015), while the per capita consumption of

red meat in Iran is about 12.5 kg (FAO,

2016). In recent years, the sharp increase in

the price of red meat in Iran has caused a

major part of the vulnerable group of the

society to reduce their consumption of this

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1754

kind of meat and to consume other kinds of

meat (including chicken meat) as a

substitute for it (Cheraghi and Gholipoor,

2010).

The uncertainty and price fluctuations of

red meat and inputs have led to the reduction

in domestic production and increase in

imports (Alijani and Saboohi, 2009).

According to FAO report (2015) and Iran

Chamber of Commerce, Industries, Mines

and Agriculture (2016), imports has always

had a larger share than exports of red meat,

and the trade balance of this product has

been negative in the last decade. Therefore,

Iran has no significant export market share

of this product. Based upon the Iran

Ministry of Agriculture Jihad (2007), one of

the most important factors affecting the

price increase of red meat is the livestock

feed. An average value of 3 billion dollar

livestock input is imported by Iran every

year (Iran Feed Industry Association, 2017).

The most important imported livestock feeds

are corn, soybean meal, and barley. Corn is

the first imported product of Iran and is one

of the main items of livestock feed that, due

to low domestic production, a significant

amount of it is imported each year

(Ghasemi, 2016). Given the goals of

reducing imports and increasing the

production of red meat in Iran at 2025

horizon, recognition of production structure

and the cost structure of beef cattle farms

seems necessary to allocate more investment

in this sector.

In 2015, the total annual amount of red

meat production in Iran was 806 thousand

tons, and Khorasan Razavi Province, with

71 thousand tons, ranked the first (Iran

Ministry of Agriculture Jihad, 2015). In the

study area of this research, Mashhad, there

are 94 beef cattle fattening farms with

operation license. They raise more than

53,600 heads of cattle and calves. The total

amount of beef produced in this region is

about 4,000 tons (Agricultural Jihad

Organization of Khorasan Razavi Province,

2017).

Beef producers often blame low farm

prices and consumers blame high retail

prices (Hosseini and Shahbazi, 2010). Due

to the shortage and high cost of feeds in

recent years, the high cost of livestock

fattening has caused some beef cattle farms

in Iran to be bankrupt, which eventually led

to a reduction in the welfare of producers

and consumers. To overcome this problem,

understanding cost structure and trying to

reduce consumer price through reducing

production costs seems necessary. Thus, this

research aimed to examine the cost structure

of red meat production in Mashhad region

through the estimation of input cost share

equations. Considering the significant

impact of the presence or absence of

economy of scale on production costs,

investigation of the economy of scale was

another goal of this research. In order to

achieve these goals, the short-run Translog

cost was used.

The study of cost structure, the economy

of scale, and the estimation of substitution

elasticity were first proposed by Christensen

et al. (1973) and Berndt and Wood (1975)

and further developed by Griffin and

Gregory (1976). Later, several researchers

used the cost function approach to analyze

the structure of production in different

sectors of the economy. In the following,

some of the studies that have investigated

the cost structure of agricultural production

are mentioned.

Grisley and Gitu (1985) investigated the

cost structure of Turkey production in the

Mid-Atlantic region using Translog variable

cost function and found that both the own-

price and cross-price elasticities of input

demands were inelastic. Glass and Mckillop

(1989) studied the structure of Northern

Ireland agriculture using a two-product,

four-input Translog cost function and

showed that the demand for livestock feed,

fertilizer, and seeds were inelastic, whereas,

the demand for labor was elastic.

Guttormsen (2002) examined input

substitutability in the salmon aquaculture

industry of Norway by estimating a Translog

cost function. The results showed that there

was no substitution between inputs. Feed

cost had also the largest share in total

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1755

production costs. Gervais et al. (2006)

estimated economy of scale for three

Canadian manufacturing agro-food sectors

including meat, bakery, and dairy using a

Translog cost function. They found that

there was evidence of economies of scale in

meat and bakery industries; however, there

was no economy of scale in the dairy

industry. Ansari and Salami (2007) studied

the economy of scale in the shrimp farming

industry in Iran using a Translog cost

function and found that there was an

increasing return to scale in this industry.

Kumar et al. (2010) estimated the input

demand functions of some Indian

agricultural products. The results indicated

that demand for modern inputs was more

elastic. Rahmani and Ghaderzadeh (2010)

estimated the cost function of poultry meat

in Sanandaj city and found that feed cost had

the largest share of total variable costs.

Kavoosi Kalashami et al. (2016) evaluated

production structure of warm water fish

farms in Guilan Province by estimating a

Translog cost function and showed that all

inputs were substitutes. Tsakiridis et al.

(2016) investigated feed substitution and

economy of scale in Irish beef production

systems. They estimated a short-run

Translog cost with panel data by using the

3SLS method. The results indicated that all

inputs were substitutes and there were

economies of scale in production systems.

Ozer and Top (2017) estimated demand for

inputs in silkworm production of Turkey and

found that the lowest price elasticity of

demand was related to the mulberry leaves

and the highest price elasticity was related to

the transportation costs. Ejimakor et al.

(2017) studied the production structure of

agricultural sector in the southeastern United

States using a Translog cost function and

concluded that the demand for labor and

other intermediary inputs were inelastic,

while, the demand for chemicals was elastic.

Together, these studies provide important

insights into the production structure of

agricultural products. However, few studies

have been conducted on the cost structure of

red meat production and most of the

research has been carried out on dairy

farming. In recent years, given the rising

production costs of beef cattle fattening, an

unprecedented rise in the price of red meat

and, consequently, the decline in consumer

demand for this product and the bankruptcy

of some of these farms, the future of this

industry in Iran is in danger. Therefore, it is

necessary to determine the most important

factors affecting beef cattle farming total

costs. This could help to adopt government

policies to reduce production costs of these

farms. Therefore, the results of this research

could be useful in this regard. Another

advantage of this research is that all the

variables that affect the total cost of beef

cattle farming have been used. The

population of this research includes all

industrial beef cattle fattening farms in

Mashhad. According to available statistics,

there are currently 94 industrial beef cattle

fattening farms in this city, of which 60

farms have been selected as samples using

Cochran formula. Data was collected from

the selected beef cattle producers in 2017 by

conducting the interview using structured

questionnaires.

MATERIALS AND METHODS

Duality theory is an appropriate tool for

estimating production functions (Diewert,

2018). According to this theory, production

structure of an industry can be examined by

both the production function and the cost

function approaches (Shefard, 1970). Since

all information about the production

structure could be recovered from the cost

function (Kavoi et al., 2010), using the cost

function instead of the production function

to investigating the cost structure of an

industry has several benefits. First, there is

no need to impose homogeneity of degree 1

on the production function to derive

equations. The cost function is

homogeneous of degree 1 in input prices,

such that doubling the prices will double the

cost without having any effect on the input

price ratio (Binswanger, 1974; Antle and

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1756

KQPQPQPQPQKP

PPPPKPPPPPKP

PPPPPPQKPP

PPQKPPPPTC

QKTQTEQEFQFLQLTTK

KEEKTEETFFKTFFTEFFELLK

TLLTELLEFLLFQQKKTTTEEE

FFFLLLQKTTEEFFLL

lnlnlnlnlnlnlnlnlnlnlnln

lnlnlnlnlnlnlnlnlnlnlnln

lnlnlnlnlnln2)(ln2

12)(ln2

12)(ln2

12)(ln2

1

2)(ln2

12)(ln2

1lnlnlnlnlnlnln 0

(6)

Aitah, 1983). Second, in the cost function

approach, instead of input quantities that are

not suitable exogenous variables at the farm

level, the input prices are used as

independent variables, because managers are

considering exogenous input prices to make

the decision to use inputs. Third, the use of

production function to estimate substitution

elasticity and the price elasticity of inputs

require that the matrix of coefficients of

production function be inverted, which leads

to an increase in estimation errors. However,

in the cost function approach, there is no

need to invert the matrix of coefficients

(Binswanger, 1974). Fourth, in estimating

the production function, there is a high

multicollinearity between independent

variables, but this problem does not arise in

the cost function (Stier, 1985).

The general form of the cost function

could be shown as follows:

),...,( 21 QPPPcC n (1)

Where, C shows the total Cost, Pi is the

Price of input i (assuming that there are n

inputs), and Q is the production Quantity.

Using the cost function to estimate the

production function parameters requires

considering of a particular functional form.

Among the flexible functional forms used in

agricultural economics studies, the Translog,

the generalized quadratic, and the

Generalized Leontief could are widely used

(Marcin, 1991; Guttormsen, 2002).

The Translog function is the most

common form of flexible functions used in

production research. One of the features of

this function in literature is that it does not

impose any prior restrictions on substitution

relationships between inputs (Christensen et

al., 1973). This function is most often used

by the researchers because of its flexibility

and variability in elasticities and returns to

scale (Sadoulet and Janvry, 1995;

MacDonald et al., 2000). The single-output

Translog cost function could be considered

as Equation (2) (Christensen and Greene,

1976; Boluk and Koc, 2010):

)ln()ln()ln()ln(2

1

)ln()(ln2

1)ln(ln 2

0

QPPP

PQQC

i

i

iqji

i j

ij

i

i

iqqq

(2)

Since the Translog cost function is linear

homogenous relative to the input prices,

input cost shares should be homogeneous of

degree 1 relative to the input prices. The

conditions of linear homogeneity of input

prices and symmetry that guarantees a well-

behaved cost function is done by applying

restrictions (3) and (4) on parameters (Boluk

and Koc, 2010).

i i j

ij

j

jiij

i

iq

i

i

0

;0;1

(3)

jiij (4)

In accordance with Shepard’s Lemma,

conditional input demand functions could be

derived by partially differentiating the cost

function with respect to the input prices

(Banda and Verdugo, 2007):

1

lnlnln

ln

j

iqjijiii

i

i

i

i QPC

XP

P

C

C

P

P

CS

(5)

Where, Si shows the cost Share of input i

and Xi is the demand quantity of input i.

Given

11

i

ii XP

, so

.11

i

iS

In this

research, a short-run Translog cost function

for beef cattle fattening farms of Mashhad is

considered as Equation (6). Also, the input

cost share equations are presented in the

form of Equation (7):

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1757

Table 1. Descriptive statistics of selected variables.

Variable Description Mean Std dev Min Max

Q Production Quantity

(kg per month) 1719.071 1878/118 125 7375

TC Total Cost

(Toman per month) 31500000 27900000 4880167 111000000

PF Feed Price

(Toman per kilogram) 1010.245 129.723 512.555 1330.039

PE Energy Price

(Toman per head of beef) 1083.243 600.95 455 4550

PT Veterinary and pharmaceutical

service Price

(Toman per head of beef)

21795.3 20805.24 2507.14 109417

PL Labor Price

(Toman-person per month) 1153167 190605.6 780000 2100000

K Capital

(Toman in the period of fattening) 30800000 25200000 4999995 100000000

SL Labor cost Share 0.074 0.0067

SF Feed cost Share 0.845 0.013

SE Energy cost Share 0.0035 0.0002

ST Veterinary and pharmaceutical

service cost Share 0.077 0.011

QKPPPPS

QKPPPPS

QKPPPPS

QKPPPPS

TQTKTTTETEFTFLTLTT

EQEKTETEEEFEFLELEE

FQFKTFTEFEFFFLFLFF

LQLKTLTELEFLFLLLLL

lnlnlnlnlnln

lnlnlnlnlnln

lnlnlnlnlnln

lnlnlnlnlnln

(7) Selected variables of this research include

production quantity, total cost, labor price,

veterinary and pharmaceutical service price,

feed price, energy price, and capital as a

quasi-fixed input. The description of these

variables is presented in Table 1.

Parameters Estimation Method

In Translog cost function literature, the most

popular estimation method is Iterated

Seemingly Unrelated Regression (ISUR) of

Zellner (1963). By using this method, cost

function and cost share equations could be

estimated simultaneously and more

efficiently than Ordinary Least Squares

(OLS) method. Application of this method

requires the estimation of the covariance

matrix for each equation that increases the

variability of the estimator sampling and

provides estimates that are numerically

closer to the maximum likelihood estimator

(Berndt, 1991; Greer, 2012). Since the sum

of the cost shares is equal to 1, there will be

only N-1 linearly independent cost share

equations (Banda and Verdugo, 2007). On

the other hand, in each equation, the sum of

the components of disturbances must be

equal to 0, which means that the covariance

matrix of disturbances is singular.

Accordingly, one cost share equation should

be deleted and its parameters through

homogeneity condition need to be

determined. The problem that arises is the

estimated value of parameters, which may

change with respect to the equation that was

deleted. For this reason, iterated seemingly

unrelated regression method is used, so that

the estimates are not sensitive to the

equation that was deleted. Also, to retain the

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1758

condition of linear homogeneity, cost

function and cost share equations should be

normalized by dividing input prices into the

input prices whose equation was deleted

from the system (Barten, 1969; Bauer et al.,

1987; Greer, 2012).

Economy of Scale

As noted by Hanoch (1975), the economy of

scale should be measured along the

expansion path, in which each point relates

to a minimum total cost and input prices are

constant. The economy of scale is defined

as: the equivalent increase in total cost due

to the equivalent increase in output, holding

input prices fixed (Filippini, 2001). The

economy of scale can be calculated from the

Equation (8):

Q

CSCE

ln

ln1

(8)

Where, SCE shows the economy of scale

and Q

C

ln

ln

is the cost elasticity. The

positive value of SCE indicates that there is

an increasing return to scale and negative

value of it indicates that there is a decreasing

return to scale (Grisley and Gitu, 1985).

Price Elasticity and Substitution Elasticity

Based on Brown and Christensen (1980),

own-price and cross-price elasticities of

input demands are calculated from

Equations (9) and (10):

niforS

SS

SS

SS

P

X

i

iiii

ii

iii

i

iiii

i

i

ii

,...,1

;ln

ln

2

2

2

(9)

jibutnjiforSS

SS

SS

SS

Pi

X

ji

jiij

ij

ijj

i

jiiji

ij

,...,1,

;ln

ln

(10)

Where, ii and ij indicate own-price and

cross-price elasticities and ii and ij

are

parameters. iS and jS

are the cost Share of

inputs i and j, which are computed as the

means of the independent variables. ij and

ii are also the Allen-Uzawa partial

elasticities of substitution (Grisley and Gitu,

1985). Allen-Uzawa elasticities evaluate the

change in demand quantity of each input

relative to a change in other input prices,

which is weighted by the cost share of the

input whose price has changed (Magnani

and Prentice, 2006), such that whatever the

cost share of the input for which price has

changed is higher, other inputs are a better

substitute or complement to it.

RESULTS AND DISCUSSION

The statistical descriptions of selected

variables are presented in Table 1. This table

shows that the monthly average beef

production of fattening farms is about 1,719

kg. Also, feed with 84.5% of total costs has

the largest cost share among the production

inputs. This finding supports the previous

research conducted by Guttormsen (2002)

and Rahmani and Ghaderzadeh (2010). This

result indicates the high importance of this

input in beef production. The lowest cost

share (0.35%) is related to the energy. It

should be noted that the price of feed (PF) is

derived from the weighted average price of

corn, straw, alfalfa, and concentrate inputs.

Also, the Price of Energy (PE) is derived

from the weighted average price of fuel and

electricity inputs.

The Translog cost function along with

input cost share equations was estimated by

imposing the constraints of symmetry and

homogeneity using Stata 14 software (For

more information about estimating cost

function in Stata, see Kumbhakar et al.

(2015)). For this purpose, cost function and

cost share equations were normalized by

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1759

Table 2. Translog cost function and cost share equations (ISUR method).

t-Statistic Coefficient Parameter t-Statistic Coefficient Parameter

-0.10 -0.00035 LT 1.59 21.545

0

-2.93 ***-0.0022 LE 2.48 **0.306

L

-2.35 **-0.019 LK 3.41 ***0.901

F

-4.16 ***-0.032 LQ -1.22 -0.223 T

-8.51 ***-0.068 FT 2.23 **0.015

E

-0.78 -0.0006 FE -0.25 -0.492

K

0.56 0.012 FK -3.04 ***-3.926 Q

1.70 *0.033 FQ 3.82 ***0.046 LL

0.87 0.0002 TE 7.18 ***0.112

FF

0.47 0.007 TK 9.63 ***0.068

TT

-0.07 -0.00095 TQ 5.95 ***0.0026 EE

0.8 0.00037 EK -0.55 -0.083

KK

-1.04 -0.0004 EQ -1.41 -0.177 QQ

2.92 ***0.347 KQ -3.50 ***-0.043 LF

R2= 0.91 Cost function

R2= 0.65 SL

R2= 0.40 SF

R2= 0.61 ST

*,

** and ***: Show that the coefficients are statistically significant at 10, 5 and 1 percent levels.

dividing the input prices into the energy

price that has the lowest cost share and cost

share equation of this input was deleted from

the system. After estimating the model, the

coefficients of the energy input were

calculated through homogeneity condition.

The results are presented in Table 2. These

results show that the coefficient of

determination (R2) of the cost function is

0.91, which means 91 percent of the

variation in production cost of beef cattle

fattening farms could be explained by the

variation in variables of labor price, feed

price, energy price, veterinary and

pharmaceutical service price, and capital.

The coefficient of determination for labor

cost share, feed cost share, and veterinary

and pharmaceutical service cost share

equations are 0.65, 0.40, and 0.61,

respectively. Glass and Mckillop (1989)

have also argued that the Translog models

yield relatively poor fits for the cost share

equations. The coefficients of the cost

function do not have an important economic

interpretation, but they are used in

estimating the price elasticity of input

demands and elasticities of input

substitution. Before estimating the

elasticities, the results of some goodness-of-

fit tests are discussed in the followings.

In order to test the linear homogeneity,

once input cost shares were estimated using

ordinary least square regression, without

imposing any constraints on parameters and,

again, these equations were estimated by

imposing homogeneity constraint on

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1760

Table 3. Results of the Wald test for homogeneity.

Probability Statistic Wald Equation

0.46 0.769 SL

0.24 1.39 SF

0.47 0.522 SE

0.09 2.60 ST

Table 4. Results of the Autocorrelation Tests.

Harvey LM Probability Harvey LM statistic Durbin-Watson statistic Equation

0.72 0.126 2.06 Cost function

0.65 0.195 1.88 SL

0.75 0.099 2.06 SF

0.87 0.024 1.94 SE

0.65 0.195 2.09 ST

0.446 (0.97)a Harvey overall system LM statistic

12.40 (0.71)a Guilkey overall system LM statistic

parameters. Then, these restrictions were

tested by the Wald test. The null hypothesis

of this test is the homogeneity of the cost

function relative to the input prices. Table 3

shows that the null hypotheses of

homogeneity are confirmed in all equations

at 5% significance level.

To test autocorrelation in cost function and

input cost shares, we used Durbin-Watson,

Harvey LM, and overall system

autocorrelation tests (For more information

about these tests, see Griffiths et al. (1985)).

The null hypothesis of these tests is that

there is no autocorrelation. As can be seen

from the Table 4, based upon Harvey LM

and Durbin-Watson tests, there is no

autocorrelation in cost function and cost

share equations. Also, based upon Harvey

overall system LM and Guilkey overall

system LM tests, there is no overall

autocorrelation in the system.

In order to test heteroscedasticity in

equations used, Engle's ARCH LM and

Hall-Pagan LM tests and Jarqu-Bera test

were applied to test normality [for more

information about these tests, see Shehata

(2011)]. The results of Table 5 indicate that

the null hypothesis of homoscedasticity is

confirmed for all equations at a 5%

significance level. Also, the results of the

Jarque-Bera test shows that the null

hypothesis of normality of disturbances is

confirmed for all equations at 5%

significance level.

After estimating the cost function and

carrying out the goodness-of-fit tests, the

cost elasticity and economy of scale were

calculated. The results are reported in Table

6. The average value of the cost elasticity is

0.465, which indicates that with the increase

of a unit of production quantity, the total

cost will increase only by 0.465. The

economies of scale for all observations have

been positive and the average of these values

is equal to 0.535, which is significantly

different from zero. It means that there are

economies of scale for all selected beef

cattle fattening farms. In other words, all of

these farms benefit an increasing return to

scale, and as production quantity increases,

costs will decrease. This finding is

consistent with that of Grisley and Gitu

(1985), Ansari and Salami (2007), and

Tsakiridis et al. (2016). The results of the

estimating own-price and cross-price

elasticities of demands are reported in Table

7. The table illustrates that the value of all

elasticities is less than 1, which indicates

that the demand for all inputs is inelastic.

Therefore, a change in input price has a

relatively small effect on the quantity of the

input demand. These results match those

observed in Grisley and Gitu (1985),

Rahmani and Ghaderzadeh (2010),

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1761

Table 5. Results of the Heteroscedasticity and Normality tests.

Normality test Heteroscedasticity tests

Equation Probability

JB

statistic Probability

Hall-Pagan LM

statistic Probability

Engle's ARCH

LM statistic

0.163 3.63 0.80 0.058 0.65 0.193 Cost

function

0.518 1.32 0.97 0.0013 0.96 0.0017 SL

0.268 2.63 0.29 1.10 0.73 0.110 SF

0.283 2.52 0.55 0.35 0.97 0.0009 SE

0.22 2.96 0.102 2.66 0.69 0.156 ST

Table 6. Estimation of cost elasticity and economy of scale.

Std dev Mean

0.189 0.465 Cost elasticity

0.189 0.535 Economy of scale

Table 7. Estimation of Own and cross price elasticities of input demands.a

Veterinary Energy Feed Labor Input *0.072

(0.04)

***0.026-

(0.009)

0.263

(0.162)

*0.304-

(0.163) Labor

-0.0034

(0.009)

***0.0023

(0.0008)

**0.022-

(0.01)

**0.023

(0.01) Feed

**0.134

(0.05)

-0.025

(0.114)

***0.676

(0.2)

***0.0554-

(0.2) Energy

-0.051

(0.09)

**0.006

(0.0025)

-0.037

(0.103)

0.07

(0.042) Veterinary

a The numbers in the parenthesis represent the standard error.

Tsakiridis et al. (2016), and Zhang and

Alston (2018).

Price elasticity of demand for all inputs

has the correct and expected negative sign,

which suggests that with the increase in the

prices, the demand for them will decrease.

Own-price elasticity of feed demand is

significant at 5% level and own-price

elasticity of labor demand is significant at

10% level, while the price elasticities of

energy and veterinary and pharmaceutical

service is not significant. Also, the results

show that the demand for feed is perfectly

inelastic, indicating a small variation in the

quantity of demand relative to a variation in

the price of this input. Therefore, feed is the

most important input of beef production.

The results of cross-price elasticities imply

that labor and energy inputs are

complementary inputs, suggesting that an

increase in labor price will decrease the use

of energy and vice versa. It is due to the fact

that labor and machinery are used together

and machinery consumes fuel and

electricity. This result supports the earlier

research by Nozari et al. (2013), Kanaani

and Ghaderzadeh (2016), and Zhang and

Alston (2018). On the other hand, labor and

veterinary and pharmaceutical service are

substitute inputs, which means that an

increase in veterinary and pharmaceutical

service price will increase the labor demand

in order to clean livestock environment to

protect them from illnesses. Also, there is a

substitution relationship between feed and

labor; since the use of more labor reduces

the loss of feed. These results corroborate

the findings of Ahmad et al. (1993) and

Shahbazi (2016). In addition, feed and

energy are substitutes. Moreover, energy and

veterinary and pharmaceutical service are

substitute inputs. Therefore, with the rising

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1762

Table 8. Estimation of Allen-Uzawa substitution elasticities of inputs.a

Veterinary Energy Feed Labor Input

0.935

(0.579)

***7.42-

(2.7)

0.311

(0.191)

*4.1-

(2.2) Labor

-0.043

(0.122)

***0.8

(0.23)

-0.026

(0.021) symmetry Feed

**1.74

(0.742)

*71.4-

(36.7) symmetry symmetry Energy

-0.662

(1.18) symmetry symmetry symmetry Veterinary

a The numbers in the parenthesis represent the standard error.

veterinary and pharmaceutical service price,

the producer uses more machinery to keep

clean the livestock environment. This

finding is in line with the results of the study

of Nozari et al. (2013). It should be noted

that there is weak substitute or

complementary relationship between inputs

because all of the cross-price elasticities are

less than 1. Hence, it can be concluded that a

change in the price of an input would not

remarkably change the demand for substitute

or complement of it. The results of estimating Allen-Uzawa

substitution elasticities are reported in Table

8. These results are consistent with the

estimated price elasticities of input demands.

Table 8 shows that there is a significant

complementary relationship between labor

and energy. This result is consistent with

Nozari et al. (2013) finding. In addition,

there is a significant substitute relationship

between feed and energy and between

energy and veterinary and pharmaceutical

service. These results are similar to those

reported by Ollinger et al. (2000). Other

Allen-Uzawa elasticities were not

statistically significant.

CONCLUSIONS

Understanding the input demand and cost

structure in beef cattle industry is essential

for evaluating the impacts of change in

government policies such as support

programs. The price elasticity of demand for

inputs is an important parameter in

quantifying the effects of these programs.

The empirical results of this study showed

that there were increasing returns to scale for

all selected farms. That means significant

cost reduction could be achieved with

increasing output level. Therefore, the

managers of these fattening farms can

increase profitability by increasing the scale

of the farms and production quantity. The

major constraint for beef cattle producers to

increase scale of the farms is shortage of

capital. Accordingly, facilitating access to

finance for beef cattle producers by giving

bank loans with low interest rates could be

helpful to expand their scale of production.

The results of the estimated elasticities

indicated that there were weak

complementary or substitute relationship

between inputs, which means that it is not

easy to substitute or complement among

inputs. According to the results of this

research, the most important factor of beef

production in the selected farms was feed,

for which the demand is inelastic and the

possibility of substituting it with other inputs

is also very weak. Therefore, an increase in

the price of this input can directly affect the

welfare of beef producers and consumers.

Thus, the adoption of policies by the

government, including subsidies for feeding

cattle and increasing the import of this input

in an effort to make livestock feeds

affordable for producers can reduce the

production cost and prevent beef prices from

rising. Furthermore, considering that the

largest share in total production costs is

related to feed cost, providing training

courses for beef cattle producers to inform

them about the optimal use of feed can be

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1763

effective in reducing production costs.

REFERENCES

1. Agricultural Jihad Organization of Khorasan

Razavi Province. 2017. Agriculture Sector

Statistical Journal.

2. Ahmad, M., Bravo-Ureta, B. and

Wackernagel, R. 1993. Factor Substitution

in Dairy Farming: A Comparison of Allen

and Morishima Elasticities. Selected Paper

at the AAEA-CAEA Annual Meeting, August

1-4, Orlando, Florida.

3. Alijani, F. and Saboohi, M. 2009. Measuring

Market Power and Cost Efficiency of Meat

and Beef in Iran. J. Agri. Econ. Res.,

1(2):77-90. (in Persian)

4. Ansari, V. and Salami, H. 2007. Scale

Economies in Shrimp Farming Industry in

Iran. J. Agri. Econ., 1 (3):35-51. (in

Persian).

5. Antle, J. M. and Aitah, A. S. 1983. Rice

Technology, Farmer Rationality, and

Agricultural Policy in Egypt. Am. J. Agric.

Econ., 65:667-674.

6. Banda, H. S. and Verdugo, L. E. B. 2007.

Translog Cost Functions: An Application for

Mexican Manufacturing. Banco de Mexico

Documentos de Investigation Working

papers.

7. Barten, A. P. 1969. Maximum Likelihood

Estimation of a Complete System of

Demand Equations. Eur. Econ. Rev., 1(1): 7-

73.

8. Bauer, P. W., Ferrier, G. and Lovell, C. A.

K. 1987. A Technique for Estimating a Cost

System That Allows for Inefficiency (No.

8704). Federal Reserve Bank of Cleveland.

9. Berndt, E. R. and Wood, D. O. 1975.

Technology, Prices, and the Derived

Demand for Energy. Rev. Econ. Stat. 259-

268.

10. Berndt, E. 1991. The Practice of

Econometrics, Classic and Contemporary.

Addison Wesley Publishing Co., Reading,

MA.

11. Binswanger, H. P. 1974. A Cost Function

Approach to the Measurement of Elasticities

of Factor Demand and Elasticities of

Substitution. Am. J. Agric. Econ., (56): 377-

386.

12. Boluk, G. and Koc, A. A. 2010. Electricity

Demand of Manufacturing Sector in Turkey:

A Translog Cost Approach. Ener.

Econ., 32(3): 609-615.

13. Brown, R. S. and Christensen, L. R. 1980.

Estimating Elasticities of Substitution in a

Model of Partial Static Equilibrium: An

Application to US Agriculture, 1947-

1979. Paper presented at the Conference

on the Economics of Substitution in

Production. Florida, USA. 14. Cheraghi, D. and Gholipoor, S. 2010. A

Review of the Main Challenges of Red Meat

in Iran. J. Com. Rev., 8(41):89-110. (in

Persian)

15. Christensen, L. R. and Greene, W. H. 1976.

Economies of Scale in US Electric Power

Generation. J. Polit. Econ., 84(4, Part 1):

655-676.

16. Christensen, L. R., Jorgenson, D. W. and

Lau, L. J. 1973. Transcendental Logarithmic

Production Frontiers. Rev. Econ. Stat., 55:

28-45.

17. Diewert, E. 2018. Duality in

Production (No. 2018-2). Vancouver School

of Economics.

18. Ejimakor, G., Quaicoe, O. and Asiseh, F.

2017. Agricultural Factor Use and

Substitution in the South-Eastern United

States. Stud. Agri. Econ., 119(3): 156-159.

19. FAO. 2015. Food Outlook. Biannual Report

on Global Food Markets.

20. FAO. 2015. Livestock Commodities.

Available Online at:

http://www.fao.org/docrep/005/y4252e/y425

2e05b.htm#TopOfPage

21. FAO. 2016. FAO Regional Office for Near

East and North Africa. Available Online at:

http://www.fao.org/neareast/news/view/en/c/

430391/

22. Fatemi Amin, S. R. and Mortezaei, A. 2013.

Food Supply Chain Strategic Plan. Jihad

University Press, Shahid Beheshti

University.

23. Filippini, M. 2001. Economies of Scale in

the Swiss Nursing Home Industry. Appl.

Econ. Lett., 8(1): 43-46.

24. Gervais, J. P., Bonroy, O. and Couture, S.

2006. Economies of Scale in the Canadian

Food Processing Industry. MPRA Paper No.

64. Available Online at: https://mpra.ub.uni-

muenchen.de/64/

25. Ghasemi, A. 2016. Deputy of Economic

Affairs. Look at the Market of Livestock and

Poultry Inputs. Deputy of Economic

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1764

Affairs, Production Research and

Policies Office. 26. Glass, J. C. and McKillop, D. G. 1989. A

Multi-Product Multi‐Input Cost Function

Analysis of Northern Ireland Agriculture,

1955–85. J. Agric. Econ., 40(1): 57-70.

27. Greer, M. 2012. Electricity Marginal Cost

Pricing: Applications in Eliciting Demand

Responses. Elsevier.

28. Griffiths, W. E., Judge, G. G., Hill, R. C.,

Lütkepohl, H. and Lee, T. C. 1985. The

Theory and Practice of Econometrics.

Wiley, 494 PP.

29. Griffin, J. M. and Gregory, P. R. 1976. An

Intercountry Translog Model of Energy

Substitution Responses. Am. Econ.

Rev., 66(5): 845-857.

30. Grisley, W. and Gitu, K. W. 1985. A

Translog Cost Analysis of Turkey

Production in the Mid-Atlantic Region. J.

Agric. Appl. Econ., 17(1): 151-158.

31. Grunert, K. G. 2006. Future Trends and

Consumer Lifestyles with Regard to Meat

Consumption. Meat Sci., 74:149-160.

32. Guttormsen, A. G. 2002. Input Factor

Substitutability in Salmon Aquaculture.

Mar. Resour. Econ., 17(2): 91-102.

33. Hanoch, G. 1975. The Elasticity of Scale

and the Shape of Average Costs. Am. Econ.

Rev., 65(3): 492-497.

34. Hosseini, S. S. and Shahbazi, H. 2010. A

Model of Iran's Farm-Retail Marketing

Margin for Beef. J. Agr. Sci. Tech., 12(3):

255-264.

35. Iran Chamber of Commerce, Industries,

Mines and Agriculture. 2016. Export and

Import Statistics. Available Online at:

http://www.tccim.ir/

36. Iran Feed Industry Association. 2017.

Available Online at:

http://www.irfia.ir/view.aspx?t=10296.

37. Iran Ministry of Agriculture Jihad. 2007.

Deputy of Animal Affairs. Available Online

at: http://www.agri-jahad.org.

38. Iran Ministry of Agriculture Jihad 2015.

Animal Production Improvement

Organization. Available Online at:

http://www.kermanshah.maj.ir/HomePage.as

px?TabID=5055&Site=kermanshahsite&La

ng=fa-IR

39. Iran Ministry of Agriculture Jihad. 2015.

Department of Animal Production.

Available Online at: http://dla.agri-

jahad.ir/Portal/Home/Default.aspx?Category

ID=e1a7847e-0ac7-42f5-9ba4-16ffc6fb25a8

40. Kanaani, F. and Ghaderzadeh, H. 2016.

Economic Analysis of the Production

Structure of Fish Farms in Kurdistan State.

J. Agri. Econ. Res., 8(29):1-20. (in Persian).

41. Kavoi, M. M., Hoag, D. L. and Pritchett, J.

2010. Economic Performance of Exotic

Dairy Cattle under Smallholder Conditions

in the Marginal Zones of Kenya Using Three

Analytical Approaches. Agrekon, 49(1): 56-

79.

42. Kavoosi Kalashami, M. Gharib Parsa, A.

and Allahyari, M. S. 2016. Evaluating

Production Characteristics and Cost

Structure of Warm Water Fisheries Farm in

Guilan Province. J. Aqua. Dev., 1(4):117-

132. (in Persian).

43. Kumar, P., Shinoj, P., Raju, S. S., Kumar,

A., Rich, K. M. and Msangi, S. 2010. Factor

Demand, Output Supply Elasticities and

Supply Projections for Major Crops of

India. Agri. Econ. Res. Rev., 23(1).

44. Kumbhakar, S. C., Wang, H. and

Horncastle, A. P. 2015. A Practitioner's

Guide to Stochastic Frontier Analysis Using

Stata. Cambridge University Press.

45. MacDonald, J. M., Ollinger, M. E., Nelson,

K. E. and Handy, C. R. 2000. Consolidation

in US Meat Packing. Department of

Agriculture and principal agency of the

Federal Statistical System of United States.

Economic Research Service. No. 785, US

46. Magnani, E. and Prentice, D. 2006.

Unionization and Input Flexibility in US

Manufacturing, 1973-1996. Indus. Lab. Rel.

Rev., 59(3):386-407.

47. Marcin, T. C. 1991. Cost Function Approach

for Estimating Derived Demand for

Composite Wood Products. In Symposium

on Systems Analysis in Forest Resources,

Charleston, SC.

48. Nozari, N., Ghaderzadeh, H. and Mirzaee,

K. 2013. A Survey on Cost Structure of

Poultry Farms: Case Study of Sanandaj and

Kamyaran Cities, Iran. J. Anim. Sci. Res.,

92(2): 83-98. (in Persian).

49. Ollinger, M., MacDonald, J. M. and

Madison, M. 2000. Structural Change in US

Chicken and Turkey Slaughter. Economic

Research Service, US Department of

Agriculture, Washington DC.

50. Ozer, O. O. and Top, B. T. 2017. Demand

for Inputs in Silkworm Production: The

Case of Turkey. Rev. Bras. Zootec., 46(12):

917-923.

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

Cost Structure of Beef Cattle Fattening Farms ___________________________________

1765

51. Rahimi Baigi, S. Kohansal, M. R. and

Dourandish, A. 2014. Prediction of Meat

Demand in Iran for Urban Areas Using

Genetic Algorithm. J. Agri. Econ., 8(3): 49-

64. (in Persian)

52. Rahmani, Z. and Ghaderzadeh, H. 2010.

Estimation of Cost Function of Broiler

Farms in Sanandaj Township. J. Agri. Econ.

Res., 6 (21): 67-80. (in Persian).

53. Sadoulet, E. and De Janvry, A. 1995.

Quantitative Development Policy Analysis.

Baltimore, John Hopkins University Press.

54. Shahbazi, A. 2016. Studying the Production

Structure of Poultry Units in Khouzestan

Province: Case Study of Masjed Soleyman

Town. Master Dissertation, Urmia

University, Iran. (in Persian)

55. Shefard, R. W. 1970. Theory of Cost and

Production Function. Princeton University

Press, New Jersey.

56. Shehata, E. 2011. LMHREG3: Stata Module

to Compute Overall System

Heteroscedasticity Tests After (3SLS-

SURE) Regressions, Agricultural

Economics Institute, Egypt.

57. Stier, J. C. 1985. Implications of Factor

Substitution, Economies of Scale, and

Technological Change for the Cost of

Production in the United States Pulp and

Paper Industry. J. For. Sci., 31(4): 803-812.

58. Tsakiridis, A., Hanrahan, K., Breen, J.,

Wallace, M. and Cathal, O. D. 2016. Feed

Substitution and Economies of Scale in Irish

Beef Production Systems. In 149th EAAE

Seminar, Structural Change in Agri-food

Chains: New Relations between Farm

Sector, Food Industry and Retail Sector

October 27-28, Rennes, France (No.

244769). European Association of

Agricultural Economics.

59. Zellner, A. 1963. Estimators for Seemingly

Unrelated Regression Equations: Some

Exact Finite Sample Results. J. Am. Stat.

Assoc., 58(304): 977-992.

60. Zhang, W. and Alston, J. M. 2018. The

Demand for Inputs and Technical Change in

the US Dairy Manufacturing Industry. Agric.

Resour. Econ. Rev., 47(3):1-35.

های مقیاس واحدهای پرواربندی گوساله گوشتی در ارزیابی ساختار هزینه و صرفه

شهر مشهد

پ. علیزاده، ح. محمدی، ن. شاهنوشی، س. سقائیان، و ع. پویا

چکیده

در ایران پرورش دام و به تبع آن، افزایش شدید قیمت گوشت قرمزهای باالی های اخیر هزینهدر سال

کنندگان به مصرف گوشت مرغ به عنوان موجب کاهش تقاضای این محصول گردیده است و مصرف

اند. این امر موجب کاهش انگیزه تولید شده و با تعطیلی برخی واحدهای جایگزینی برای آن روی آورده

کنندگان این محصول با خطر جدی مواجه شده است. برای کنندگان و مصرفپرواربندی دام، رفاه تولید

رفع این مشکل، شناخت ساختار هزینه و تالش جهت کاهش قیمت تمام شده این محصول از طریق

رسد. از این رو، هدف مطالعه حاضر، ارزیابی ساختار هزینه و های تولید ضروری به نظر میکاهش هزینه

باشد. برای این منظور تابع دهای پرواربندی گوساله گوشتی در شهر مشهد میهای مقیاس واحصرفه

های های تولید با استفاده از روش رگرسیونهزینه کوتاه مدت ترانسلوگ به همراه توابع سهم هزینه نهاده

و از های مورد استفاده در این مطالعه به صورت پیمایشی اند. دادهبه ظاهر نامرتبط تکراری برآورد شده

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0

______________________________________________________________________ Alizadeh et al.

1766

طریق مصاحبه حضوری با مدیران واحدهای پرواربندی گوساله گوشتی شهر مشهد با استفاده از

گردآوری شده است. نتایج نشان دهنده وجود بازدهی صعودی 6931پرسشنامه ساختار یافته در سال

ها قیمت آنها نسبت به تغییرات چنین تقاضای نهادهنسبت به مقیاس برای تمامی واحدها بوده است. هم

های تولید برقرار باشد. از سوی دیگر رابطه مکملی و جانشینی بسیار ضعیفی بین نهادهکشش میکامالً بی

ترین عامل تعیین کننده تولید گوشت در واحدهای مورد بررسی باشد. بر اساس نتایج این مطالعه، مهممی

ها نیز امکان جانشینی آن با سایر نهاده ناپذیر بوده ونهاده خوراک دام بوده است که تقاضای آن کشش

هایی از سوی دولت از جمله پرداخت یارانه خوراک دام به باشد. بنابراین اتخاذ سیاستبسیار اندک می

های پرورش دام تواند باعث کاهش هزینهمی تولیدکنندگان این محصول و افزایش واردات این نهاده

ند.شده و از افزایش قیمت گوشت جلوگیری ک

Dow

nloa

ded

from

jast

.mod

ares

.ac.

ir at

5:3

2 IR

ST

on

Wed

nesd

ay D

ecem

ber

30th

202

0